1. Field of the Invention
The present invention relates to a method of changing a bit rate of a compressed image bit stream, and more particularly, to an apparatus and a method of changing a bit rate of a compressed image bit stream by changing a codeword so as to easily change the bit rate and reduce a changing loss.
2. Description of the Related Art
In line with the recent development of Internet-related technologies, images as well as text and voice data are quickly distributed throughout the Internet. Particularly, an industry which provides image services through the Internet is rapidly growing.
In general, a service provider transmits image data from a source, through diverse networks, to a destination. A destination party negotiates with the service provider for the Quality of Service (QoS), considering the capacity and price of a receiving terminal, network situations, and so on. One of the leading QoS parameters used is a bit rate.
After the negotiation, if a compressed image bit stream which has been already stored does not match the agreed on QoS, the service provider has to change the compressed image bit stream to satisfy the agreed on QoS. Generally, the agreed on QoS is lower than the QoS of the stored compressed image bit stream.
A Personal Video Recorder (PVR) function, including simultaneous playback and recording, is also in demand by users of televisions and image receiving apparatuses for the Internet. With the PVR function, a receiving apparatus temporarily stores a received compressed image bit stream in a Hard Disc Drive (HDD). Since the users want to store as many programs as possible in the HDD, changing the bit rate of the compressed image bit stream is also needed with respect to the PVR function.
Methods which lower a bit rate of a compressed image bit stream without changing a syntax of the compressed image bit stream are broadly divided into a transform coefficient re-quantization method and a transform coefficient selection method.
In a transform coefficient re-quantization method, a transform coefficient, for example, a discrete coefficient transform DCT coefficient, is re-quantized using a new quantization process so as to have the bit rate satisfy a desired output bit rate. Generally, this method is performed through a variable length decoder, a dequantizer, a re-quantizer, and a variable length encoder, and has a complicated structure.
In a transform coefficient selection method, part of the transform coefficients are selected so as to reconstruct the output bit rate. The transform coefficient selection method has a simpler structure than the transform coefficient re-quantization method.
FIG. 1 shows a block diagram of the structure of a conventional apparatus which changes a bit rate using a known transform coefficient selection method. The apparatus comprises a coefficient selection adjuster 104 which selects part of decoded transform coefficients that were decoded through a variable length decoder 102. Here, coefficients to be selected are determined by a rule of the coefficient selection adjuster 104.
FIGS. 2A and 2B show a zonal masking type and a breaker pointer type of the conventional transform coefficient selection method.
As shown in FIG. 2A, using a zonal mask Bz, which is determined with respect to an output bit rate, transform coefficients outside the zonal mask Bz are discarded while the transform coefficients inside the mask are selected.
As shown in FIG. 2B, using a break pointer Bp, which is determined with respect to an output bit rate, transform coefficients which are scanned after the break pointer Bp are discarded while transform coefficients which are scanned prior to the break pointer Bp are selected.
Since the transform coefficient re-quantization method is performed through a variable length decoder, dequantizer, re-quantizer, and a variable length encoder, the transform coefficient re-quantization method has a structure more complicated than that of the transform coefficient selection method using the variable length decoder, the coefficient selector, and the coefficient selection adjuster. However, the transform coefficient selection method discards part of the transform coefficients, which results in a picture quality that is inferior to that of the transform coefficient re-quantization method.